Intel(R) Math Kernel Library for Deep Neural Networks (Intel(R) MKL-DNN)  1.0.4
Performance library for Deep Learning
Enumerations
Queries

Enumerations

enum  mkldnn_query_t {
  mkldnn_query_undef = 0, mkldnn_query_engine, mkldnn_query_primitive_kind, mkldnn_query_num_of_inputs_s32,
  mkldnn_query_num_of_outputs_s32, mkldnn_query_time_estimate_f64, mkldnn_query_memory_consumption_s64, mkldnn_query_scratchpad_engine,
  mkldnn_query_impl_info_str, mkldnn_query_some_d = 64, mkldnn_query_op_d, mkldnn_query_convolution_d,
  mkldnn_query_deconvolution_d, mkldnn_query_shuffle_d, mkldnn_query_eltwise_d, mkldnn_query_softmax_d,
  mkldnn_query_pooling_d, mkldnn_query_lrn_d, mkldnn_query_batch_normalization_d, mkldnn_query_inner_product_d,
  mkldnn_query_rnn_d, mkldnn_query_gemm_d, mkldnn_query_some_md = 128, mkldnn_query_src_md,
  mkldnn_query_diff_src_md, mkldnn_query_weights_md, mkldnn_query_diff_weights_md, mkldnn_query_dst_md,
  mkldnn_query_diff_dst_md, mkldnn_query_workspace_md, mkldnn_query_scratchpad_md
}
 Primitive descriptor query specification. More...
 

Detailed Description

Enumeration Type Documentation

◆ mkldnn_query_t

Primitive descriptor query specification.

For generic function mkldnn_primitive_desc_query(), the type of result must agree with the queried argument. The correspondence table:

Query | type of result

mkldnn_query_engine | mkldnn_engine_t * mkldnn_query_scratchpad_engine | mkldnn_engine_t * mkldnn_query_primitive_kind | mkldnn_primitive_kind_t * *_s32 | int * *_s64 | mkldnn_dim_t * (same as int64_t *) *_f64 | double * *_str | const char ** mkldnn_query_op_d | const_mkldnn_op_desc_t * *_md | const mkldnn_memory_desc_t ** *_${op}_d | const mkldnn_${op}_desc_t ** *_pd | const_mkldnn_primitive_desc_t *

Note
Rule of thumb: all opaque types and structures are returned by reference. All numbers are returned by value.
Warning
All returned references point to constant objects and are valid only during the lifetime of the queried primitive descriptor. Returned objects must not be destroyed by the user. If you need to keep the object longer than the lifetime of the queried primitive descriptor, use mkldnn_primitive_desc_clone() to make a copy.
Enumerator
mkldnn_query_undef 

no query

mkldnn_query_engine 

execution engine

mkldnn_query_primitive_kind 

primitive kind

mkldnn_query_num_of_inputs_s32 

number of inputs expected

mkldnn_query_num_of_outputs_s32 

number of outputs expected

mkldnn_query_time_estimate_f64 

runtime estimation (seconds)

mkldnn_query_memory_consumption_s64 

memory consumption – extra (scratch) memory, additional to all inputs and outputs memory (bytes)

mkldnn_query_scratchpad_engine 

scratchpad engine – engine to be used for creating scratchpad memory

mkldnn_query_impl_info_str 

implementation name

mkldnn_query_some_d 

stub

mkldnn_query_op_d 

op descriptor

mkldnn_query_convolution_d 

convolution descriptor

mkldnn_query_deconvolution_d 

deconvolution descriptor

mkldnn_query_shuffle_d 

shuffle descriptor

mkldnn_query_eltwise_d 

eltwise descriptor

mkldnn_query_softmax_d 

softmax descriptor

mkldnn_query_pooling_d 

pooling descriptor

mkldnn_query_lrn_d 

lrn descriptor

mkldnn_query_batch_normalization_d 

batch normalization descriptor

mkldnn_query_inner_product_d 

inner product descriptor

mkldnn_query_rnn_d 

rnn descriptor

mkldnn_query_gemm_d 

GEMM descriptor.

mkldnn_query_some_md 

stub

mkldnn_query_src_md 

source memory desc

mkldnn_query_diff_src_md 

source gradient memory desc

mkldnn_query_weights_md 

weights memory descriptor desc

mkldnn_query_diff_weights_md 

weights grad. memory desc

mkldnn_query_dst_md 

destination memory desc

mkldnn_query_diff_dst_md 

destination grad. memory desc

mkldnn_query_workspace_md 

workspace memory desc

mkldnn_query_scratchpad_md 

scratchpad memory desc